From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:58107) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Y3004-0003Zs-RR for qemu-devel@nongnu.org; Mon, 22 Dec 2014 05:14:09 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Y3000-0007db-0e for qemu-devel@nongnu.org; Mon, 22 Dec 2014 05:14:04 -0500 Received: from szxga01-in.huawei.com ([119.145.14.64]:5666) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Y2zzz-0007cw-CE for qemu-devel@nongnu.org; Mon, 22 Dec 2014 05:13:59 -0500 Message-ID: <5497EED0.7070205@huawei.com> Date: Mon, 22 Dec 2014 18:13:36 +0800 From: zhanghailiang MIME-Version: 1.0 References: <1417849159-6568-1-git-send-email-zhang.zhanghailiang@huawei.com> <1417849159-6568-4-git-send-email-zhang.zhanghailiang@huawei.com> <20141221211734.15420.72276@loki> In-Reply-To: <20141221211734.15420.72276@loki> Content-Type: text/plain; charset="utf-8"; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH RFC for-2.3 3/6] qga: implement qmp_guest_get_memory_blocks() for Linux with sysfs List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Michael Roth , qemu-devel@nongnu.org Cc: hangaohuai@huawei.com, peter.huangpeng@huawei.com On 2014/12/22 5:17, Michael Roth wrote: > Quoting zhanghailiang (2014-12-06 00:59:16) >> We can get guest's memory block information by using command >> "guest-get-memory-blocks", the returned value contains a list of memory block >> info, such as phys_index, online state, can-offline info. >> >> Signed-off-by: zhanghailiang >> --- >> qga/commands-posix.c | 66 +++++++++++++++++++++++++++++++++++++++++++++++----- >> 1 file changed, 60 insertions(+), 6 deletions(-) >> >> diff --git a/qga/commands-posix.c b/qga/commands-posix.c >> index 8917dca..d3f7d4f 100644 >> --- a/qga/commands-posix.c >> +++ b/qga/commands-posix.c >> @@ -2005,6 +2005,60 @@ static void transfer_memory_block(GuestMemoryBlock *mem_blk, bool sys2memblk, >> g_assert(res == 0); >> } >> >> +GuestMemoryBlockList *qmp_guest_get_memory_blocks(Error **errp) >> +{ >> + GuestMemoryBlockList *head, **link; >> + Error *local_err = NULL; >> + struct dirent *de; >> + DIR *dp; >> + >> + head = NULL; >> + link = &head; >> + >> + dp = opendir("/sys/devices/system/memory/"); >> + if (!dp) { >> + error_setg_errno(errp, errno, "Can't open directory" >> + "\"/sys/devices/system/memory/\"\n"); >> + return NULL; >> + } >> + >> + /* Note: the phys_index of memory block may be discontinuous, >> + * this is because a memblk is the unit of the Sparse Memory design, which >> + * allows discontinuous memory ranges (ex. NUMA), so here we should >> + * traverse the memory block directory. >> + */ >> + while ((de = readdir(dp)) != NULL) { >> + GuestMemoryBlock *mem_blk; >> + GuestMemoryBlockList *entry; >> + >> + if ((strncmp(de->d_name, "memory", 6) != 0) || >> + !(de->d_type & DT_DIR)) { >> + continue; >> + } >> + >> + mem_blk = g_malloc0(sizeof *mem_blk); >> + /* The d_name is "memoryXXX", phys_index is block id, same as XXX */ >> + mem_blk->phys_index = strtoul(&de->d_name[6], NULL, 10); >> + mem_blk->has_can_offline = true; /* lolspeak ftw */ > > My initial thought as well :P > Ha, yes. ;) >> + transfer_memory_block(mem_blk, true, &local_err); >> + >> + entry = g_malloc0(sizeof *entry); >> + entry->value = mem_blk; >> + >> + *link = entry; >> + link = &entry->next; >> + } >> + >> + if (local_err == NULL) { >> + /* there's no guest with zero memroy blocks */ > > typo > >> + g_assert(head != NULL); > > As commented on in earlier patches, guest errors shouldnt crash the guest agent > if it can be avoided. > OK. >> + return head; >> + } >> + >> + qapi_free_GuestMemoryBlockList(head); >> + error_propagate(errp, local_err); >> + return NULL; >> +} >> #else /* defined(__linux__) */ >> >> void qmp_guest_suspend_disk(Error **errp) >> @@ -2040,6 +2094,12 @@ int64_t qmp_guest_set_vcpus(GuestLogicalProcessorList *vcpus, Error **errp) >> return -1; >> } >> >> +GuestMemoryBlockList *qmp_guest_get_memory_blocks(Error **errp) >> +{ >> + error_set(errp, QERR_UNSUPPORTED); >> + return NULL; >> +} >> + >> #endif >> >> #if !defined(CONFIG_FSFREEZE) >> @@ -2126,12 +2186,6 @@ GList *ga_command_blacklist_init(GList *blacklist) >> return blacklist; >> } >> >> -GuestMemoryBlockList *qmp_guest_get_memory_blocks(Error **errp) >> -{ >> - error_set(errp, QERR_UNSUPPORTED); >> - return NULL; >> -} >> - > > Looks like some unecessary code movement made it's way into the patch. Please > squash this change into original patch > OK, will fix that. Thanks. >> int64_t qmp_guest_set_memory_blocks(GuestMemoryBlockList *mem_blks, >> Error **errp) >> { >> -- >> 1.7.12.4 > > > . >